The Pride Project Team describe a state of the art review and user requirements, including user scenarios and technical, commercial and legal implementation factors. PRIDE is an interoperability project developing software to create an international distributed directory of library services and resources. The directory will store information about Patrons and Services and will be used to provide a range of interlibrary loan, personalised alerting, information retrieval and payment functions.
People and Resources Identification for Distributed Environments (PRIDE) [1] is a two year project with funding from the European Commission TAP Libraries Programme [2], which started in June 1998. The project will develop a distributed directory service supporting a range of library services and users. It will demonstrate solutions to various technical and managerial problems that arise when libraries attempt to provide their users with efficient one-stop access to dispersed organised information in printed and electronic forms.
PRIDE will enable access via a single point to a global range of information resources in a way that supports a large number of services and functions. It will provide:
 PRIDE project teams in the UK, Ireland, France, Germany, Hungary and Australia have completed a systematic specification of “PRIDE Requirements and Success Factors”, and are about to move into the design and implementation phase of the project, with a view to demonstrating services from the latter half of 1999. This article seeks to explain the objectives of the project and some of our findings to date. It is not too late for others to influence progress of the project, and, through this paper, the project hopes to attract interest, critical comment and particularly additional sponsors, who will help create an effective and far-reaching brokerage service for the global information community. |
It has been found useful to consider global information infrastructures as landscapes. Users may navigate from one place to many others with greater or lesser ease. Most users will follow well-trodden paths, but they also tend to mould the landscape themselves through the ways in which they interact with ‘features’ of the landscape, such as resources and resource collections.
Human knowledge develops through making and strengthening some linkages at the expense of others. The global information infrastructure will develop in the same way. PRIDE will help libraries and existing resource sharing agencies to develop the ‘synapses’ of the global electronic brain, but commercial players are likely to become increasingly important.
In many cases, libraries will be able to expose their whole user populations to publishers’ products through middleware, with advantages for all parties. Libraries will themselves therefore come to behave more like publishers, selectively adding value in various ways for a range of target audiences and marketing themselves appropriately for each one. Libraries are in a strong position to provide high quality networked services and to improve the experiences of many users of the Internet and conventional libraries.
Libraries must help users make information connections
Certain characteristics of the Information Society affect PRIDE:
Libraries risk losing users to services better able to add value for users.
PRIDE will oil the information works. Directories can support a wider range of uses of existing library services for more people, without requiring wholesale replacement of systems. The PRIDE directory architecture will evolve according to need, although in the two year project we will only be able to begin to demonstrate the possibilities.
The scope of its wide-ranging consortium partnerships will ensure not only that a range of services becomes available quickly in several European countries, but that PRIDE leverages some of the world’s most advanced practical thinking about interlibrary loans and document delivery systems, in Australia.
The project has divided its staff resources into three groups for the purpose of defining PRIDE requirements and success factors, as described below.
The first group, with most input from the London and South Eastern Library Region (LASER), which is co-ordinating the project and Australian partner Macquarie University, Sydney, looked at user requirements and scenarios.
Directories will have wide application, but PRIDE will concentrate on the most important features lacking in the current environment, which are solutions for making distributed Patron information and distributed Service information available.
In general, at present library users:
The library itself has three main options for handling users:
1. Treat them all alike. This creates no barriers to use and suits users well, except where personalised or customised services need to be provided, but is likely to mean that resources which can be provided are severely limited. Little or no administration overheads may accompany low use of the network. Also, libraries do not provide any added value in terms of administrative convenience for users. Furthermore, even a specialist library, which focuses on a single user community and provides the same service for everyone, may want to restrict access to members only, which places them in the next category.
2. Treat them as groups. The commonest way to provide access to academic materials, this option relies on identifying the user within a group somehow – usually by using an IP address or shared username/password or some combination of these. Administration overhead is determined by the number of groups. This option typically works best within a campus type institution where there is some control over IP addresses and users. Group information cannot easily be shared across institutions without the use of intermediary directories. Users may benefit from group membership of services, but they may find that the number of groups they belong to transparently is confusing, so they may be unaware of all the services to which they are entitled.
3. Treat them as individuals. This gives information providers the maximum flexibility to group users, exchange rights information, develop personalised and customised services. However, it is unlikely that any other service will be able to share this information without intensive proprietary effort. Services following this option can deliver much added value to users, but a user will either have to restrict the number of such services used or will have to deal with a large number of separate systems offering different passwords, commands and search languages.
Libraries therefore:
The availability of distributed Patron information (not restricted to the local library automation system) will facilitate a range of services that can be applied to users regardless of their library services affiliation (membership). These include:
The PRIDE project seeks:
The availability of distributed Service information will facilitate all the services currently envisaged for the global library space. For example:
Information stored in PRIDE directories will include both service and collection descriptions [3], and not just those describing traditional library catalogue materials.
Some of these services are essential if the concept of the virtual, distributed or federated library is to become a business reality, whilst others add significant value to the opportunities. They are complementary to core services (such as Search, Locate, Request and Deliver) and must inter-operate seamlessly with them.
The PRIDE project seeks:
Beyond the library world, generic and focused services will be developed for the home, the office and the public place by major corporations, offering searching and payment options, support for hobbies and interests, edutainment and lifelong learning. To guarantee a place in the public eye, library services must therefore be able to inter-operate with those services in terms of
The PRIDE application therefore will provide distributed Patron and Service / Collection information in a manner that is capable of integration with the broader offerings in the wider digital services marketplace.
Given the inevitable distribution, specialisation & also competition between these services, a number of issues become critical. There must not be a restrictive view of a single 'top node' (root) in any directory structure. Standards implementation profiles need to be established to allow directory service providers to inter-operate and to contribute to a common downstream view. Inheritance and local/domain specialisation must be implemented.
The PRIDE project has elected to establish directory services for each member state (UK, Ireland, France and Germany) in such a manner that they can seamlessly interoperate to provide cross-border services - and indeed global services as indicated by the participation of an Eastern European (Hungary) and a Pacific Rim (Australia) state in the project.
PRIDE will build on the work of the EU funded LIRN project [5], which produced a directory of business library service information.
Finally the PRIDE demonstrator will incorporate the dimension of 'people' services operating beyond the world of libraries and networked information brokerage:
These links to the wider world of service developments will be achieved through the contributions of sponsoring partners and through direct exploitation of relevant Framework Programme demonstrators, and interested parties are encouraged to contact the project early in 1999.
PRIDE is developing a range of application scenarios which will be supported by the distributed directory service.
 Figure 1: Large-Scale Distributed Searching |
The user submits a query to the PRIDE server to ascertain which databases hold information on the required topic. The PRIDE server responds with a list of potential targets based on its knowledge of their contents and provides the client with contact information for the targets. This contact information includes address details as well as semantic inter-operability information gleaned from its Explain service (e.g. query attributes supported, schema and record syntax details etc). The client then dynamically re-configures itself in line with the available services from the target databases. The query is then entered against the list of supported attributes for the targets and issued to them in parallel.
Alternatively, the user can issue the initial search to the PRIDE server and the server will forward the query to the appropriate databases based on its knowledge of their content.
Distributed database searching is now becoming possible through the use of parallel Z39.50 [7] clients. However, the initial problem of finding out which databases should be searched is as yet unsolved. PRIDE aims to provide this initial referral capability and integrate it with distributed searching to provide an end-to-end information discovery service. Additionally, as more Z39.50 servers are becoming available, the problems of semantic inter-operability are becoming more apparent and the need for dynamic interface configuration is now well recognised. Through its Explain service the PRIDE server will provide dynamic inter-operability information to connected clients.
 Figure 2: Distributed Authentication for Document Delivery |
After locating an item of interest (using the above scenario) the user issues a document request either through Z39.50 item order or through an ISO ILL request message to the supplier. The supplier must establish the authenticity of the user and obtain payment information. To do this the supplier contacts the PRIDE directory service, which establishes a 'trust path' to the user and obtains account information for that user.
This is essential for all service providers who wish to provide large scale commercial services. It should be noted that at present large scale document delivery services do not deal directly with the public as a result of this problem (for example BLDSC only deals with libraries who operate on behalf of members of the public).
Without such a service in place, end users will never be able to deal directly with document suppliers.
The Australian partner in the PRIDE project has contributed details of how such a Directory service would be applied to the well advanced Australian interlending and resource sharing developments, giving extremely valuable advance information of priorities. The PRIDE website will list other relevant projects
The second project team consists of academic and commercial partners from University College, Dublin, the UK Office for Library and Information Networking (UKOLN) at the University of Bath and Fretwell-Downing Informatics. The PRIDE directory services will essentially provide accessible ‘forward’ information about people, rights and resource collections and have the capability to support directly or indirectly the translation of queries and responses from one protocol to another, so that PRIDE can act as a brokerage service. It is important to realise that we do not yet fully understand all the services which will be provided by PRIDE or which will be enabled through others’ use of PRIDE services. Nor do we understand what will be the final optimal implementation infrastructure for a network of PRIDE and related directories. The scope of our completed technology survey demonstrates a huge range of potentially relevant standards, techniques and software. The main purpose of the survey is to help partners design the architecture of the PRIDE service and select the technologies to use. We had two goals in collecting the information:
The availability of such a comprehensive collection of technical analysis and advice will inform the managerial decisions about to be taken regarding the demonstrator network, and should reduce the risk of obsolescence.
We have examined a wide range of user interfaces, query languages, search engines. Directory service models and directory standards studied include X.500, Lightweight Directory Access Protocol (LDAP), Microsoft Active Directory (MAD), WHOIS++. We have also examined Corba Trader standards from the Object Management Group (OMG), and the Domain Name System (DNS).
Search and Retrieve systems have been analysed, as the work force of Information Retrieval (IR). They provide access to various document databases, performing selection functions according to the user requests. There are three important characteristics that distinguish different Search and Retrieve systems:
Search techniques are differentiated by the types of search requests they accept; for example, a hierarchical system will present to the user a set of alternatives to choose from, while a keyword-based search system will ask for keywords to look for.
We have looked at types of search request; the Z39.50 protocol; the present and future of HyperText Transfer Protocol (HTTP); the ILL protocol (ISO 10160/1), developed in the context of managing the entire document request lifecycle within a distributed environment; Electronic Data Interchange (EDI) including EDIFACT (Electronic Data Interchange For Administration, Commerce And Transport, ISO 9735); CORBA Object Query Service, which is a general service for querying networked information services alongside CORBA facilities for Trader (distributed directories), Interface Repositories and Type Management.
In addition, we have collected information about harvesting and Web crawlers, meta-information, security and payment systems from a technical point of view. It is likely that we shall choose either to make PRIDE behave transparently to these technologies, or to select a small number of them for implementation in the working demonstrator. We shall concentrate on those relating to authorisation, collection description and payment, as these are least well understood in the libraries sector.
A third study has been undertaken, by LITC (Library Information Technology Centre) at South Bank University, London, of the commercial and legal issues which we either face already or which may arise during the project. This sought to provide a brief management overview of a wide range of issues which may affect the design or operation of a distributed PRIDE directory service. It highlights potential problem areas and makes recommendations to the other project partners on functionality and project management.
Complex areas reviewed from a PRIDE perspective included copyright law, international data protection legislation, self-regulation options. We also considered commercial issues such as the potential need to make international electronic contracts or pay taxes on transactions. The need to comply with licenses and regulations through differing service levels for users is raised, along with questions of marketing and how to position the PRIDE service so that users are aware of it and can make use of it.
PRIDE services will abide by the law. Unfortunately, we do not know what this is for many likely scenarios, particularly in the international context. Non-library interests drive many new laws, and some efforts are being made internationally by libraries, especially in the area of copyright. During the short timescales for the PRIDE project, we are inevitably forced to adopt various levels of self-regulation. We are seeking to develop consensus solutions, and to follow best practice through case studies and liaison with those working in these fields.
One example of self regulation will be the handling of personal information.
Personal information is widely available because it is collected by numerous organisations such as schools, universities, business (employers and telephone companies) and membership records. While the protection of privacy on the Internet is in its infancy, PRIDE will adopt a self-regulation policy,This policy may involve the selection of those sites for inclusion in the directory that adopt data protection principles of disclosure and informed consent or adopt the use of a branded on-line seal or trustmark to signify compliance with personal information policy.
This policy is already being used by organisations like CommerceNet [8] and the Electronic Frontier Foundation (EFF) [9]. Sites that display the trustmark have formally agreed to adhere to privacy principles, disclose their information gathering and dissemination practices, and submit to a comprehensive assurance process.
The main industry principles include removing certain categories of non-essential, personally identifiable data from information products and services such as financial or medical records, removing the records of individuals under the age of eighteen from locator service products.
As PRIDE moves towards customised services for users, the need for handling contracts with end users will inevitably arise, whether these are made with PRIDE itself or with providers of PRIDE linked services.
The development of model contract clauses to guarantee the protection of personal data is one possibility for dealing with the EU principle on transfer of personal data overseas but the problem of enforcing such a contract to protect the data subject is still being considered.
The International Chamber of Commerce [10] has written a model contract for cybertrading in relation to data protection. Under the terms of the model clauses companies outside the European Union would undertake to give personal data emanating from within the EU the same level of protection as the EU’s data protection directive. The clauses also provide for legal remedies if consumers believe their legal rights on privacy have been breached.
All electronic contracts try to impose certain terms and conditions such as use of services, maximum liabilities, payment methods, disclaimers and various other points. It is assumed that PRIDE will be using electronic contracts for the purpose of agreement between users.
This raises questions of liability and applicable law. For example, web-wrap or click-wrap agreements are web page accept buttons which oblige the user to accept the standard terms and conditions before they see the rest of the site or place a query. Although no English case law exists concerning web-wrap contracts these were upheld in the US in the case of Hotmail v Van Money Pie 1998 [11].
Under the draft US Uniform Commercial Code 2B regulations [12], contract terms must be presented at the time of access to, or immediately prior to the placing of a query on an online service. The user must also be required to take certain affirmative action to indicate acceptance, such as by clicking on an accept button and be able to cancel the query at any time before transmission. European rules on electronic commerce,have not yet been made clear, although the European Commission has proposed Directives allowing electronic contracts to be made using digital signatures [13][14][15].
From the selection of points above, it is clear that not all PRIDE users can be treated equally. PRIDE will therefore be able to support a range of service levels, based not merely on technical or geographical issues but also on commercial and legal ones.
Questions arise about who will be responsible for managing the necessary operational services, co-ordinating development activities and liasing with contracted repositories. How can we distribute management most effectively and efficiently?
In order to protect against abuse by users, PRIDE Administration will have to maintain certain records, including the identity of subscribers (for authentication and authorisation) and the terms and conditions agreed by them.
One key area we hope to explore is links to payment systems. In a global environment PRIDE will need the support of at least one bank to be able to support multiple payment models in order to do business with customers. By supporting credit card payments, for example, we shall pursue issues relating to the authentication of individuals, transactions and security, including use of public key infrastructures (PKIs), certificates, Certification Authorities (perhaps for the library community specifically, along the lines of the international IFLA interloans voucher scheme), digital signatures and digital passports.
John Kay's book (OUP, 1993) 'Foundations of Corporate Success' is recognized by Arthur Andersen Consulting staff as identifying four areas in which ventures must succeed: innovation, branding, access to strategic assets and good business architecture (the web of relationships within which the business operates).
As a whole, assessing libraries against this suggests:
An early aim could be to establish PRIDE library services as an entity which could be added to Net directories, portals and homepages of other organizations through simple links which would results from negotiations with the operators. Trials and pilots combined with careful measurement of results would then guide future marketing efforts. This strategy might then result in some restructuring of supplier services to better match demand.
In the traditional library world, we have learnt that not all users want to do all the work of information (and book) management themselves. Libraries can respond collectively more quickly and effectively to publication growth, market forces, new technologies than individuals can themselves, although their needs have always driven service development. Because of libraries, new publications, new ways of organising information and new ideas spread more quickly, in ways that are still poorly understood because they are so pervasive.
Effectively, what PRIDE is doing is allowing managed and organised changes in the clustering of resources and users to be made easier. Services should then improve. These kinds of changes will happen without the implementation of directories, but their use will give libraries a higher profile in the wider information society, leveraging existing investment in the skills of library and information staff.
End users may well play a larger role in the development of the network than hitherto, through their decisions to use certain features of services, to release personal information or to pay for information services.
Library managers will have to reconsider what services to provide, where to focus staff support and what level of audience to aim for, as well as how to fund the services. PRIDE will help lubricate change, but both library staff and their users will be able to steer paths more easily, without needing a detailed understanding of information retrieval, metadata or directory protocols. In the end, few are concerned with the efficiency of ‘engines’ such as PRIDE themselves. Although directories will be ubiquitous, most people will be too busy creating and exploring the collective mind of the planet.
 Figure 3: PRIDE Project Team |
Back row (left to right):
Peter Tolgyesi, Peter Cheng, Peter Smith, Nikita Schmidt,
Andy Powell, Alexey Marinichev, Geoff Butters, Robin Yeates, Peter
Brophy.
Front row (left to right):
Kerry Blinco, Paul Harvey, Andrew Colleran, Eckhard Blume.
If you have any comments on this article, please contact the editors (exploit-editor@ukoln.ac.uk).
The PRIDE project team
Liaison contact: Robin Yeates, LITC, UK
Andrew Colleran, Quercus Information Limited, Oxford, UK
Email: <Andrew.Colleran@quercus.co.uk>
URL: <http://lirn.viscount.org.uk/pride/>
For citation purposes:
The PRIDE project team, "Oiling the Works: the PRIDE Project Develops an Information
Brokerage Service," Exploit Interactive, issue 1, 10 April 1999
URL: <http://www.exploit-lib.org/issue1/pride/>
|
Issue Home | Editorial | Features | Regular Columns | News and Events | Et cetera | ||
|
| ||
| Go to Top |
A UKOLN Service. Contact Us. Copyright © 1999-2006 
|
Last Updated: 9 May 1999 |